We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Predicting Drift in the North Pacific Garbage Patch: A Benchmark of Data-Driven and Physics-Based Models
AI summary Read the abstract
Researchers compared computer models that predict how floating ocean debris drifts in the North Pacific Garbage Patch, testing both physics-based methods and data-driven approaches trained on ocean current data. They checked these predictions against real GPS tracking data from drifting buoys and tagged debris. Data-driven models performed better than physics-based ones for short-term forecasts of up to 72 hours.
The ability to accurately forecast floating debris’ short-term drift is key to efficiently removing legacy plastic pollution from the ocean. This is a challenging task, as it typically depends on integrating velocity fields, whose accurate forecast is inherently difficult. Additionally, the complex, nonlinear interplay of currents, wind and waves in the transport of material is still poorly understood. The aim of this study is to generate a comprehensive benchmark of state-of-the-art Lagrangian transport models to improve operational forecasts, with a regional focus in the North Pacific Garbage Patch. Physics-based advection models are considered, in addition to data-driven deep learning and tree-based regression approaches. These methods are applied to velocity fields generated by both neural network-based emulators and numerical ocean circulation models. The transport models' prediction skill is evaluated against several GPS drifter types and GPS-tagged debris trajectories. This analysis identifies the best-performing models and quantifies the influence of diverse geophysical fields on material transport in the region. Preliminary results indicate that data-driven approaches outperform physics-based advection models across short-term (≤72-hour) forecast horizons.
More Papers Like This
Drift Field Net: Learning Ocean Lagrangian advection fields from in-situ and satellite observations
AI summary Read the abstract
Scientists built an AI tool that predicts ocean currents more accurately than current forecasting systems, helping track where floating plastic debris drifts in the Pacific. Better predictions mean cleanup crews can target garbage patches more effectively, which matters because ocean microplastics eventually make their way into seafood and drinking water.
A Real-time Product to Help Ocean Cleanup Operations
AI summary Read the abstract
A real-time operational product was developed to simulate the drift of ocean plastic debris using ocean current models derived from satellite and drifter data, aimed at improving the efficiency of ocean cleanup operations. The tool helps predict where old, weathered plastic debris accumulates to guide search and recovery efforts.
Comparing models and observations of the surface accumulation zone of floating plastic in the North Atlantic subtropical gyre
AI summary Read the abstract
This study compared ocean circulation models to observational data on where floating plastic accumulates in the North Atlantic subtropical gyre. Improving model accuracy is important for predicting plastic concentration zones and designing effective ocean cleanup strategies.
Wind‐Based Estimations of Ocean Surface Currents From Massive Clusters of Drifters in the Gulf of Mexico
AI summary Read the abstract
Researchers used ocean surface drifter data from the Gulf of Mexico to develop models estimating surface currents from wind measurements. This type of modeling can be applied to predict how microplastic debris disperses across ocean surfaces after entering from river or coastal sources.
Pathways of marine debris derived from trajectories of Lagrangian drifters
AI summary Read the abstract
Researchers applied a probabilistic model to global satellite-tracked ocean drifter trajectories to map marine debris pathways, identifying five subtropical convergence zones maintained by Ekman currents where floating debris — including microplastics — preferentially accumulates, confirming predictions with direct ocean surface measurements.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.